Latino women are at high risk for cardiovascular (CVD) and other chronic diseases, in part, due to sedentary lifestyle, obesity, and poor fitness. Few studies assess long-term physical activity, fewer observe maintenance and none have demonstrated effective procedures for sustaining exercise. Only one study has demonstrated increased moderate exercise and none has demonstrated maintenance of physical activity among Latino women. Theoretically-based procedures (shaping, contingency management, relapse prevention, and social reinforcement) hypothesized as necessary to maintain physical activity have not yet been tested experimentally. The purpose of this study is to determine the differential effectiveness of a culturally tailored program to shape and maintain moderate intensity physical activity and to improve cardiorespiratory fitness among low SES sedentary Latino women. 210 Latinas will be assigned at random to one of 3 groups: 1 six month physical activity intervention, 1 physical activity plus maintenance intervention, and 1 safety education control. Exercise training will emphasize walking. Bilingual exercise leaders, aided by peer models from the community, will use shaping procedures to establish daily walking in participants. In the maintenance program, family incentives will reinforce transfer to community exercise and will be used to establish positive feedback loops to promote activity within the family. Peer led community exercise sessions, and community activism will also be used to establish social networks which will reinforce sustained physical activity. Program support will be faded out as naturally occurring social support for Community exercise is established. Four (baseline, post- intervention, post-maintenance, follow-up), repeated measures over 24 months will assess fitness (VO2max), physical activity (PAR), and CVD risk factors. Theoretically important mediating variables, such as self- efficacy, social support, stage of change for exercise, decisional balance, and home exercise environment will be explored. Repeated measures analyses will be used to determine significant differences among groups, time and group by time main effects. This study is the first to attempt to engineer maintenance of physical activity among minority women. Results will inform interventions targeted to Latino women and will provide a model to be used to sustain physical activity and reduce the risk of cardiovascular and other diseases. If effective, this analysis will also serve as a model for designing programs to sustain physical activity in the increasingly sedentary general population.